A method for training a neural network to derive, from an image of a camera mounted on a robot, a movement vector for the robot to insert an object into an insertion. The method includes controlling the robot to hold the object, bringing the robot into a target position in which the object is inserted in the insertion, for a plurality of positions different from the target position controlling the robot to move away from the target position to the position, taking a camera image by the camera and labelling the camera image by a movement vector to move back from the position to the target position and training the neural network using the labelled camera images.

A method for training a neural network to derive, from an image of a camera mounted on a robot, a movement vector to insert an object into an insertion. The method includes, for a plurality of positions in which the object held by the robot touches a plane in which the insertion is located controlling the robot to move to the position, taking a camera image by the camera and labelling the camera image with a movement vector between the position and the insertion in the plane and training the neural network using the labelled camera images.

An insertion quality determinator is a determinator 1 that determines a quality of insertion of an insertion component 5 inserted into a hole formed in a work object. The insertion component 5 at least includes a head having the size that is impossible to be inserted into the hole, and a pillar-shaped body that extends from the head and has the thickness that is possible to be inserted into the hole. The determinator is configured to determine the quality of insertion based on positions of given parts P1-P4 of the head of the insertion component 5 inserted into the hole, in a direction perpendicular to an extending direction of the hole.

Disclosed in the present invention is a method for plugging/unplugging a probe of a metallurgical automatic sublance. A laser distance sensor is mounted on an end-effector of a drive device; the drive device drives the laser distance sensor to scan the automatic sublance according to setting areas; position and orientation information of the automatic sublance is calculated by a calculation unit; according to the position and orientation information of the automatic sublance, a gripper on the end-effector implements a process of plugging/unplugging the probe of the automatic sublance. According to the present invention, an existing automatic sublance does not need to be improved, and only external sensors need to be used for plugging/unplugging detection, so that the plugging/unplugging process can be accurately carried out.

A press frame for a robot system includes a base, a bridge and a set of columns supporting the bridge above the base. A first robot holds a part on the base and a second robot manipulates a pressing tool to press a component into an opening. The pressing tool is backed by the bridge that opposes a reaction force resulting from pressing the component part into the part. A method of assembling components to a part by pressing the part into an opening while engaging the bridge of a reaction frame. The part is transferred to the base by a first robot that positions the part on the base. A pressing tool and a component are selected by a second robot that orients the component to be inserted in the opening. Data relating to displacement, load and time is collected by the controller.

A flexible pressing system for accepting and rejecting pressed part into a part may include a pressing apparatus configured to press components into a hole of a part and a controller programmed to receive press data for a press of at least one of the components, the press data including force, distance and time of the press, and determine whether the force is indicative of an inadequate press based on the force and distance at a specific time of the press.

An integrated control system for a flexible pressing system may include a first robot including a gripper for manipulating a part, a second robot including a pressing tool, and a controller configured to instruct the first robot to move the part into a pressing position and to instruct the second robot to concurrently ready the pressing tool for pressing.

A force detector provided between a robot arm and a robot hand detects forces Fx, Fy, and Fz. A robot controller performs a filtering process for the forces Fx and Fy by a first low-pass filter of a cutoff frequency F.sub.c1, moves the robot hand so that the forces Fx and Fy become smaller, corrects a trajectory of the robot arm, performs a filtering process for the force Fz by a second low-pass filter of a cutoff frequency F.sub.c2 having a frequency higher than the cutoff frequency F.sub.c1, performs a threshold value determination for the force Fz, and stops the movement of the robot hand when the force Fz exceeds a threshold value during a fitting operation.

A device with a processing unit and a storage device also has a base, a cover including a plurality of guiding holes, a plurality of receiving members, an installation member, a plurality of first sensing devices, and a plurality of second sensing devices. The processing unit determines by a specific one of the first sensing devices corresponding to a specific one of the receiving members that the installation member is to acquire a fastening element from the specific receiving member. Then, the processing unit determines by a specific one of the second sensing devices corresponding to a specific one of the guiding holes that the installation member is to move towards the specific guiding hole. The processing unit controls the installation member to fasten the fastening element to an electronic device through the specific guiding hole when the specific first sensing device corresponds to the specific second sensing device.

A robot for performing an assembly operation is provided. The robot comprises a processor configured to determine a control law for controlling a plurality of motors of the robot to move a robotic arm according to an original trajectory, execute a self-exploration program to produce training data indicative of a space of the original trajectory, and learn, using the training data, a non-linear compliant control law including a non-linear mapping that maps measurements of a force sensor of the robot to a direction of corrections to the original trajectory defining the control law. The processor transforms the original trajectory according to a new goal pose to produce a transformed trajectory, update the control law according to the transformed trajectory to produce the updated control law, and command the plurality of motors to control the robotic arm according to the updated control law corrected with the compliance control law.